The reader, de Man contin- ues, is led to assent to the
incompatibility
or aporia between the imagi- nation's failure and its success by "a constant, and finally bewildering alternation of the two terms, Angemessen(heit) and Unangemessen(heit), to the point where one can no longer tell them apart" (AI 90).
Paul-de-Man-Material-Events
It gets under your skin or into your nose, and "there is a certain reaction which is bound to occur.
" You sneeze or break out in a rash.
The figure is only a figure.
It compares what happens to some people in reading de Man to what happens in a certain material reac- tion to a foreign substance by a living organic body.
The figure is not innocent, however.
In comparing something seemingly "abstract," in- tentional, linguistic, or "spiritual," reading, to something material, au- tomatic, autonomic, and involuntary, something "bound to happen," that is, an allergic reaction, the question of the relation of language to "materiality" is raised.
Does any substantial connection justify the fig- ure?
This is one of the central questions in de Man's conception of a "material event.
" How can a linguistic act, such as the formulations reached by Kant's philosophic rigor, intervene in the "material" world and bring about what de Man calls "the materiality of actual history"?
2 How can writing or reading be a material event?
How can speech be an act?
As I shall show, de Man's transformation of the usual meaning of "materiality" (the transformation is itself a speech act) goes by way of a new conception of the relation of language to that reconceived materiality.
Almost any page of de Man's work, but especially the beginnings and endings of essays, contains rejections of well-established received ideas about literary study. These rejections can best be characterized as ironically and joyfully insolent or even contemptuous, as well as dis- mayingly rigorous and plausible. 3 Salient examples are the first two pages of "The Resistance to Theory" and the last three pages of "Shelley Disfigured. "4 De Man's essays have the structure he identifies in "The Concept of Irony" as "the traditional opposition between eiron and alazon, as they appear in Greek or Hellenic comedy, the smart guy and the dumb guy" (AI 165). De Man is of course the eiron, the smart guy, and all the previous experts on whatever topic or text he is discussing are the alazons, the dumb guys. 5 The received ideas he attacks, often fundamental assumptions of our profession, are charac- teristically called aberrant, deluded, or simply false. The reader can only hope or assume that "This does not, cannot, mean me! Surely I would not make such stupid mistakes. " De Man forestalls that defen- sive move, however, when he asserts, for example, in the "Kant and Schiller" essay in Aesthetic Ideology, that everyone, including himself,
however ironically, in a collective "we," is still bewitched by aesthetic ideology:
Before you either contest this [what he has been saying about Schiller's distortion of Kant], or before you not contest but agree with it and hold it against Schiller, or think that it is something we are now far beyond and that we would never in our enlightened days do--you would never make this naive confusion between the practical and the pragmatic on the one hand and the philosophical Kantian enterprise on the other-- before you decide that, don't decide too soon that you are beyond Schiller in any sense. I don't think any of us can lay this claim. Whatever writing we do, whatever way we have of talking about art, whatever way we have of teaching, whatever justification we give ourselves for teaching, whatever the standards are and the values by means of which we teach, they are more than ever and profoundly Schillerian. They come from Schiller, and not from Kant. (AI 142)
De Man goes on to make a warning that certainly applies to what has happened in his own case, in spite of the fact that he was protected by being a Sterling Professor at Yale, which is about as much security as you can get:
And if you ever try to do something in the other direction [in the direc- tion of Kant, that is, rather than Schiller] and you touch on it you'll see what will happen to you. Better be very sure, wherever you are, that your tenure is very well established, and that the institution for which you work has a very well-established reputation. Then you can take some risks without really taking many risks" (AI 142).
I have said that de Man's work is threatening to "us all" because al- most any page contains cheerfully taunting rejections, explicit or im- plicit, of "our" most basic ideological assumptions, the ones "we" most need to get on with our work, the ones the university most needs to get on with its work. His counterintuitive concept (it is not really a concept) of materiality is an example of this.
DE MAN'S MATERIALISM
The "'s" in this subhead is a double genitive, both objective and sub- jective. It names both de Man's theory of materiality and the way his own writings may show materiality at work or may be examples of materiality at work. De Man's materiality is one of the most difficult and obscure parts of his work.
Paul de Man as Allergen 185
186 J. Hillis Miller
De Man's use of the terms materiality and materialism poses several special problems, resistances to comprehension. First, one or the other word is most often introduced only briefly and elliptically. If the reader does not keep a sharp eye out for it, it appears in a given essay for an instant, for the blink of an eye, like a meteor, and then vanishes. Moreover, in these passages de Man seems to be saying exceedingly strange things, such as the assertion that materiality is not "phenome- nal. " Second, unlike "performative" and "irony" (terms not on every- one's lips and concepts that clearly need some explaining), we tend to think we already know what materiality is. It is the property possessed by these hard objects right in front of me now, impassive, impassible, resistant, not dependent on my perception for their continued exis- tence, like that stone Samuel Johnson kicked to refute Berkeley's ideal- ism: "I refute him thus [kicking the stone]. " Third, the term material- ism is extremely difficult to extricate from its associations with modern empirical science or with vulgar understandings of Marxism. Is not Marxism to be defined as "dialectical materialism"? De Man is sup- posed to be in one way or another a linguistic formalist, someone who believed, as all so-called deconstructionists are supposed to believe, that it is "all language," though the reader might remember that de Man began his higher education as a science, mathematics, and engi- neering student at the E? cole Polytechnique of the University of Brussels (1936). His professional interest in language came later. Nevertheless, for de Man to call himself a materialist, or for us to call him one, seems as absurd and counterintuitive as for de Man to call Kant and Hegel materialists or to find crucial materialist moments in their work, since everybody knows (without necessarily having read them) that they are "idealists. " Equally absurd would be to think one might find any kin- ship between de Man's thinking and Marxism, though the truth is that a deep kinship exists between de Man's work and Marx's thought in The German Ideology, as Andrzej Warminski has been demonstrating in his seminars. To show this it is necessary actually to go back and read Marx, as well as de Man, no easy tasks.
The term materiality or its cognates appears at crucial moments in de Man's work as early as a citation from Proust in "Reading (Proust)" in Allegories of Reading. What Proust calls the "symbols," in Giotto's Allegory of the Virtues and Vices at the Arena in Padua, meaning rep- resentations like the Charity that looks like a kitchen maid, are "some- thing real, actually experienced or materially handled. "6 That this pas- sage was important to de Man is indicated by the way he cites it again
at a crucial moment on the symbol in Hegel just at the end of one of his late essays, "Sign and Symbol in Hegel's Aesthetics. " This time de Man translates the phrases himself somewhat differently from the Moncrieff translation, and he cites the French original: "the symbol represented as real, as actually inflicted or materially handled [. . . (le symbole repre? sente? ) comme re? el, comme effectivement subi ou mate? riellement manie? ]" (AI 103). The terms material, materiality, and the like then appear with increasing frequency in de Man's later work. It is as though de Man had discovered in such words a way to "call" more ac- curately something he wanted performatively to name, perhaps even to invoke, that is, to "call forth": "The only word that comes to mind is that of a material vision . . . " ("Phenomenality and Materiality in Kant," in AI 82). What Michael Riffatere misses or evades in Hugo's "E? crit sur la vitre d'une fene^tre flamande" is just what the title indi- cates or names, namely, what de Man calls "the materiality of an in- scription" (RT 51). A climactic passage in Shelley's The Triumph of Life is said to stress "the literal and material aspects of language" (RR 113). "Anthropomorphism and Trope in the Lyric" ends, in a phrase I have already cited, with an appeal to "the materiality of actual history" (RR 262). A cascade of such terms punctuates the essays in Aesthetic Ideology, not only in "Phenomenality and Materiality in Kant" and in "Kant's Materialism," where "a materialism that Kant's posterity has not yet begun to face up to" (AI 89) is the focus of the argument, but also in "Sign and Symbol in Hegel's Aesthetics," where we read that "The idea, in other words, makes its sensory appearance, in Hegel, as the material inscription of names" and also in the way Hegel's "theory of the sign manifests itself materially" (AI 102, 103), and in "Kant and Schiller," where we read of the irreversible progression "from states of cognition, to something which is no longer a cognition but which is to some extent an occurrence, which has the materiality of something that actually happens, that actually occurs" and of "the materiality of the inscribed signifier in Kant" (AI 132, 134).
The reader will have seen that the term materiality and its cognates occur in three related, ultimately more or less identical, registers in de Man: the materiality of history, the materiality of inscription, and the materiality of what the eye sees prior to perception and cognition. In all three of these registers, as I shall try to show, materiality is associat- ed with notions of performative power and with what seems materiali- ty's opposite, formalism. In all three modes of materialism, the ultimate paradox, allergenic idea, or unintelligibility is the claim or insinuation
Paul de Man as Allergen 187
188 J. Hillis Miller
that materiality is not phenomenal, not open to the senses. Just what in the world could that mean?
The phrase "materiality of history" seems the easiest to understand and accept as commonsensical. Of course history is material. It means what really happened, especially as a result of human intervention (though we speak, for example, of the history of the mollusks, or of geo- logical history). History is wars, battles, the building of the pyramids, the invention of the steam engine, migrations of peoples, legislative de- cisions, diplomatic negotiations, the clearing of forests, global warm- ing, that sort of thing. De Man's materiality of history, however, is not quite like that. For him the materiality of history, properly speaking, is the result of acts of power that are punctual and momentary, since they are atemporal, noncognitive and noncognizable performative utter- ances. History is caused by language or other signs that make some- thing materially happen, and such happenings do not happen all that often. The most radical, and allergenic, counterintuitive, scandalous formulation of this is in "Kant and Schiller. " There de Man asserts that Kant's Critique of Judgment was an irreversible historical event brought about by the shift from cognitive to efficaciously performa- tive discourse in Kant's own words, whereas Schiller's ideological mis- reading of Kant and its long progeny in the nineteenth and twentieth centuries were nonevents, certainly not irreversible material events. In "Phenomenality and Materiality in Kant" de Man speaks of the crucial shift to a "formal materialism" in Kant's Critique of Judgment as "a shift from trope to performance" that is "a deep, perhaps fatal, break or discontinuity" (AI 83, 89, 79). This is the place, as he puts it in "Kant and Schiller," at which Kant "found himself by the rigor of his own discourse [the project of aesthetics as articulation of pure reason and practical reason or ethics] to break down under the power of his own critical epistemological discourse" (AI 134). This was an event, strictly speaking an irreversible historical event, "to some extent an oc- currence, which has the materiality of something that actually hap- pens, that actually occurs. And there, the thought of material occur- rence, something that occurs materially, that leaves a trace on the world, that does something to the world as such--that notion of oc- currence is not opposed in any sense to the notion of writing" (AI 132). Since the event of Kant's materialism is punctual and instanta- neous, it is in a curious sense not within time, though it has a perma- nent and irreversible effect on what we usually (mistakenly) think of as the temporality of history:
history is not thought of as a progression or a regression, but is thought of as an event, as an occurrence. There is history from the moment that words such as 'power' and 'battle' and so on emerge on the scene. At that moment things happen, there is occurrence, there is event. History is therefore not a temporal notion, it has nothing to do with temporality [there's allergenic assertion for you! ], but it is the emergence of a lan- guage of power out of a language of cognition. (AI 133)
I do not think de Man meant that the words power and battle are in themselves always historical events in the sense de Man is defining such events, but that he means the uses of such words in effective per- formative utterances are historical events. As opposed to the moment of Kant's self-undoing materialism in the third Critique, Schiller's recu- peration of Kant within aesthetic ideology and its long progeny, the procedures of which are identified in the main body of "Kant and Schiller," did not happen, were not historical events:
One could say, for example, that in the reception of Kant, in the way Kant has been read, since the third Critique--and that was an occur- rence, something happened there, something occurred [de Man's stut- tering iterations here mime the punctualities of historical events; the reader will remember that this is the transcript of an oral presentation that was not written down as such]--that in the whole reception of Kant from then until now, nothing has happened, only regression, noth- ing has happened at all. Which is another way of saying there is no his- tory . . . that reception is not historical. . . . The event, the occurrence, is resisted by reinscribing it in the cognition of tropes, and that is itself a tropological, cognitive, and not a historical move. (AI 134)7
These sternly recalcitrant statements may be more understandable and perhaps even more acceptable if we remember that Althusser, and de Man in his own way, following Marx, define ideology as having no history, as being outside history, as having no purchase on his- tory, since ideology is precisely an illusory misunderstanding of the "real conditions of existence," as Althusser put it in "Ideology and Ideological State Apparatuses,"8 or, as de Man puts this in "The Resis- tance to Theory": "What we call ideology is precisely the confusion of linguistic with natural reality, of reference with phenomenalism" (RT 11). 9 The reception of Kant by Schiller and his followers, including you and me as inheritors of aesthetic ideology, is ideological, therefore not historical.
Paul de Man as Allergen 189
190 J. Hillis Miller
We are (I am) now in a position to answer the puzzling assertions de Man makes in "The Concept of Irony. " "Irony," he says, "also very clearly has a performative function. Irony consoles and it promises and it excuses" (AI 165). What could de Man mean by saying that irony is performatively efficacious, that it promises, consoles, or excuses? If we take seriously de Man's claim later in the essay that irony is a perma- nent parabasis that radically suspends meaning by the incursion of chaos, madness, and stupidity (Friedrich Schlegel's terms) into lan- guage, then it would seem radically counterintuitive to say that irony has a successful performative function. A statement at the end of the essay is equally baffling: "Irony and history seem to be curiously linked to one another" (AI 184). If irony is permanent parabasis it would seem to have little to do with history, but to be rather the withdrawal from effective historical action. The analogy between the noncognitive aspect of irony and the noncognitive aspect of performative utterances gives the clue. Irony is perhaps the most radical example of the rupture between cognitive and performative discourses. Insofar as an utterance is performative, it is unknowable. Irony suspends cognition. It is just because irony is error, madness, and stupidity that it can be performa- tively felicitous. Promises, excuses, consolations can be performed by irony, or can be especially done by ironic utterance, just because irony is the radical suspension of cognition. Another way to put this is to say that even the most solemn performative utterances are contaminated by being possibly ironic. Jacques Derrida includes irony along with lit- erature among the parasitical presences that are possibly incorporated within any performative as a result of its intrinsic iterability.
What I have just said will also indicate the surprising and "curious" connection of irony with history. Since the materiality of history as event is generated by acts of linguistic power, that is, performative speech acts, though by no means necessarily intentional ones, irony as a form of such power or as an ingredient of any such act of power, against all our instinctive assumptions, can be said not only to prom- ise, console, and excuse, but also to generate the events that make up the materiality of history. Just as, for Derrida, the possibility of felici- tous speech acts depends on the possibility that they may be "litera- ture," so for de Man the efficacy of performative utterances, including those that generate history, depends on the possibility that they may be ironical. They may be. You cannot tell for sure.
If speech acts generating history are, strangely enough, one form of materiality or are the place where language touches materiality, leaves
a mark on it, materially handles it, the materiality of what the eye sees appears more obvious but turns out to be more difficult to grasp. Of course, we say, what the eye sees is material. That received opinion or doxa turns out, however, once again not to be quite what de Man means. What he does mean is the central argument of the two essays on Kant, "Phenomenality and Materiality in Kant" and "Kant's Materialism. " For received opinion, what we take for granted, phe- nomenality and materiality are the same thing or are two aspects of the same thing. Because something is material it is phenomenal, open to the senses. For de Man, following Kant, phenomenality and materiali- ty are not conjoined but opposed. How can this be? De Man sees in Kant's theory of the dynamic sublime two radically contradictory no- tions. On the one hand, the sublime is the moment when the imagina- tion triumphs over fear and puts all the elements of the sublime scene together, articulates them in a grand aesthetic synthesis, as tropes ar- ticulate, or as the body's limbs are articulated: "The imagination over- comes suffering, becomes apathetic, and sheds the pain of natural shock. It reconciles pleasure with pain and in so doing it articulates, as mediator, the movement of the affects with the legal, codified, formal- ized, and stable order of reason" (AI 86). In so doing, the imagination of the sublime or the sublime itself accomplishes the goal of the third Critique, which was to find a "bridge" between the first and second Critiques, between pure reason and the practical reason of moral obli- gation and choice. On the other hand, Kant's analysis of the dynamic sublime contains a moment that radically disrupts, interrupts, and sus- pends this happy articulation. Kant reaches this moment through the very rigor of his critical thinking. He proposes that the paradigmatic example of the dynamic sublime is when the overarching vault of the sky and the outstretched mirror of the sea are seen just as the eye sees them, or as the poets see them, without thought for their meaning. Seeing them as meaningful would occur, for example, when we view the sea as a reservoir of edible fish, or the sky as a producer of life-giving rain. De Man quotes section 28 of Kant's The Critique of Judgment: "we must regard it [the starry heaven], just as we see it [wie man ihn sieht], as a distant, all-embracing vault [ein weites Gewo? lbe]. . . . To find the ocean nevertheless sublime we must regard it as poets do [wie die Dichter es tun], merely by what the eye reveals [was der Augeschein zeigt]" (AI 80). De Man goes on to argue that this way of seeing is radi- cally nonphenomenal. It does not involve the mind that in its activity of perception would make sense of what is seen. It just sees what it
Paul de Man as Allergen 191
192 J. Hillis Miller
sees, in an activity of the eye operating by itself, enclosed in itself, wholly detached, disarticulated, from thinking and interpreting: "No mind is involved in the Kantian vision of ocean and heaven. . . . That is how things are to the eye, in the redundancy of their appearance to the eye and not to the mind, as in the redundant word Augenschein, . . . in which the eye, tautologically, is named twice, as eye itself and as what appears to the eye" (AI 82). De Man's name for this way of seeing is "material vision": "The only word that comes to mind is that of a material vision" (AI 82), which is another way of saying, in a paradig- matic performative speech act, "I call this 'material vision. '" The word material then appears in a cascade of phrases in the subsequent pages: "the vision is purely material"; "what we call the material aspect"; "a materialism that, in the tradition of the reception of the third Critique, is seldom or never perceived"; "If the architectonic then appears, very near the end of the analytics of the aesthetic, at the conclusion of the section on the sublime, as the material disarticulation not only of na- ture but of the body [traditional examples of the beautiful or the sub- lime], then this moment marks the undoing of the aesthetic as a valid category. The critical power of a transcendental philosophy undoes the very project of such a philosophy leaving us, certainly not with an ideology--for transcendental and ideological (metaphysical) principles are part of the same system--but with a materialism that Kant's pos- terity has not yet begun to face up to" (AI 83, 88, 89).
How could we "face up to" something that we can see but not face up to in the sense of clearly confronting it and making it intelligible to ourselves? The idea of a way of seeing that is performed by the eye alone, wholly dissociated from the mind, is, strictly speaking, unintelli- gible, since any sense we give to this Augenschein is an illicit, ideologi- cal imposition: "To the extent that any mind, that any judgment, inter- venes, it is in error" (AI 82). That is what I mean by saying that de Man's materiality is nonphenomenal, since phenomenality always in- volves, instantly, making sense or trying to make sense of what we see. This "material vision" would be pure seeing prior to any seeing as the sort of understanding that we name when we say, "I see it all now. " It would be a pre-seeing seeing, that is, something unthinkable, unknow- able, unintelligible, a tautological eye eyeing: "Realism postulates a phenomenalism of experience which is here being denied or ignored. Kant's looking at the world just as one sees it ('wie man ihn sieht') is an absolute, radical formalism that entertains no notion of reference or semiosis" (AI 128).
The idea of a materiality that would not be phenomenal does not make sense. Nevertheless, that is just what de Man affirms, most overtly and in so many words at the end of the essay on Riffaterre, "Hypogram and Inscription. " There he speaks of "the materiality (as distinct from the phenomenality) that is thus revealed [when we remember that Hugo's poem was supposed to have been written on a window pane], the unseen 'cristal' whose existence thus becomes a certain there and a certain then which can become a here and a now in the reading 'now' taking place" (RT 51). The paradox is that the window glass, figure here for the materiality of inscription, is not what the eye sees but what the eye sees through. In the Kant essays, as in "Hypogram and In- scription," the rigor of de Man's own critical thinking brings him re- peatedly, by different routes, across the border of the intelligible and into the realm of the allergenic, in this case the recognition of a materi- alism in Kant that has seldom or never been recognized in the whole distinguished tradition of Kant scholarship and so is anathema to it, just as de Man's reading of somewhat similar material moments in Hegel was anathema to the distinguished Hegel specialist Raymond Geuss. 10
The final version of materiality in de Man is the "prosaic materiali- ty of the letter" (AI 90). Just what does de Man mean by that? No one doubts that writing (and speaking too) have a material base, marks on paper or modulated waves in the air. This materiality is the benign base of the meaning, permanence, and transmissibility of language. No problem. De Man of course does not mean anything so in agreement with common sense and received opinion. When de Man calls Kant's sublime Augenschein of sky and sea a "material vision" he goes on to raise a further question that is not answered until the end of the essay: "how this materiality is then to be understood in linguistic terms is not, as yet, clearly intelligible" (AI 82). The answer is the materiality of the letter, but just what does that mean? The essay ends with an ex- planation that if not clearly intelligible, at least indicates why these "linguistic terms" must be unintelligible. The reader is given intelli- gence of unintelligibility, new news of the unknowable.
The prosaic materiality of the letter, linguistic "equivalent" of a ma- terialism of vision, has two main features. One is a disarticulation of language equaling the disarticulations of nature and the human body de Man has found in Kant's dynamic sublime: "To the dismemberment of the body corresponds a dismemberment of language, as meaning- producing tropes are replaced by the fragmentation of sentences and
Paul de Man as Allergen 193
194 J. Hillis Miller
propositions into discrete words, or the fragmentation of words into syllables or finally letters" (AI 89). Strictly speaking, as linguists, not to speak of language philosophers like Wittgenstein, have shown, words do not have meaning by themselves. They have meaning only when they are used, incorporated into sentences. To detach them from their sentences and leave them hanging there in the air or on the page, surrounded by blank paper, is the first stage in a progressive disarticu- lation of meaning that goes then to syllables and finally to letters. It is extremely difficult to see words, syllables, or letters, for example on a printed page, in this way, just as it is extremely hard to see as the eye sees. One has to be a poet, as Kant says, to do it. The mind instantly in- terprets what the eye sees, "perceives it," and gives meaning to it, just as the mind projects meaning into those mute letters on the page. It is almost impossible to see letters as just the material marks they are. Even words in a language we do not know are seen as language and not as sheer materiality. We tend to see random marks on a rock as possibly writing in an unknown language.
The other feature of the materiality of the letter stressed by de Man makes that materiality more likely to be glimpsed, in the wink of the eye, before the mind starts "reading. " This is repetition of words and word parts that calls attention to the absurd and unmotivated echoes among them at the level of syllable and letter: puns, rhymes, allitera- tions, assonances, and so on, that is, precisely those linguistic features poets especially use, "the play of the letter and of the syllable, the way of saying . . . as opposed to what is being said" (AI 89). The "persua- siveness" of the passage in Kant about the recovery of the imagination's tranquillity through material vision depends, de Man says, "on the proximity between the German words for surprise and admiration, Verwunderung and Bewunderung" (AI 89).
The reader, de Man contin- ues, is led to assent to the incompatibility or aporia between the imagi- nation's failure and its success by "a constant, and finally bewildering alternation of the two terms, Angemessen(heit) and Unangemessen(heit), to the point where one can no longer tell them apart" (AI 90). One ad- ditional example of this in de Man's essays is the cascade of words in "fall" that he finds in a passage by Kleist: Fall, Beifall, Su? ndenfall, Ru? ckfall, Einfall, Zuru? ckfall, Fa? lle: "As we know from another narra- tive text of Kleist ["On the Gradual Formation of Thoughts while Speaking"], the memorable tropes that have the most success (Beifall) occur as mere random improvisation (Einfall) at the moment when the author has completely relinquished control over his meaning and has
relapsed (Zuru? ckfall) into the extreme formalization, the mechanical predictability of grammatical declensions (Fa? lle)" (RR 290). By the time the reader gets to the end of this the root "fall" is fast becoming a mere surd, a sound emptied of meaning: "fall, fall, fall, fall. " The read- er will see that "formalism" of "formalization" names for de Man not the beautiful aesthetic formalization of the artwork, but a principle of mechanical senselessness in language that he associates with the arbi- trariness of grammar, of declensions, Fa? lle. De Man goes on to make a pun of his own. Since Falle also means trap in German, he can say that everyone falls into "the trap of an aesthetic education which inevitably confuses dismemberment of language by the power of the letter with the gracefulness of a dance. " That trap, however, is not a benign aes- theticizing of the random formalizations of language in grammar and paronomasia such as poets are known to play with. It is a mortal dan- ger, a pericolo de morte, according to the last words of the last essay in The Rhetoric of Romanticism, "the ultimate trap, as unavoidable as it is deadly" (RR 290). The reader will note that this aspect of the materi- ality of the letter tends to disappear in translation. It depends on the unique idiom, idiolect, or even "idiocy," in the etymological sense, of a certain language. Ultimately, this repetition of words and bits of words empties language of meaning and makes it mere unintelligible sound, as when the poet Tennyson, as a child, used to repeat his own name over and over, "Alfred, Alfred, Alfred," until it ceased to mean anything at all and he melted into a kind of oceanic trance. Try it with your own name, as I do here with mine: "Hillis, Hillis, Hillis, Hillis. "
De Man's formulation of this in one notable place is more prosaic. As he shows, Hegel's theory of memory as Geda? chtnis, in opposition to Erinnerung, is that it memorizes by emptying words of meaning and repeating them by rote, as pure arbitrary signs that might be in a foreign language or in no language at all:
"It is well known," says Hegel, "that one knows a text by heart [or by rote] only when one no longer associates any meaning with the words; in reciting what one thus knows by heart one necessarily drops all ac- centuation. " [I suppose Hegel means that one repeats the words mind- lessly, like a schoolchild or a robot--JHM. ] . . . The idea, in other words, makes its sensory appearance, in Hegel, as the material inscrip- tion of names. (AI 101-2)
Speaking in "Hegel on the Sublime" of Hegel's "Gesetz der A? usserlichkeit (law of exteriority)," de Man says, "Like a stutter, or a broken record,
Paul de Man as Allergen 195
196 J. Hillis Miller
it makes what it keeps repeating worthless and meaningless" (AI 116). This had already been exemplified in a truly vertiginous couple of paragraphs in "Sign and Symbol in Hegel's Aesthetics. " There de Man takes two at first innocent-enough-looking, but in fact "quite astonish- ing," sentences in Hegel's Encyclopedia: "Since language states only what is general, I cannot say what is only my opinion [so kann ich nicht sagen was ich nur meine]," and "When I say 'I,' I mean myself as this I to the exclusion of all others; but what I say, I, is precisely any- one; any I, as that which excludes all others from itself [ebenso, wenn ich sage: 'Ich,' meine ich mich als diesen alle anderen Ausschliessenden; aber was ich sage, Ich, ist eben jeder]" (AI 97, 98). The sentences themselves are bad enough in English, though worse in German (e. g. , wenn Ich sage, Ich, meine ich mich"), but by the time de Man gets through with these sentences the reader is dizzied by the repetitions, like Tennyson repeating his own first name, or as if he had been caught in a revolving door. 11 Through this dizziness the reader reaches in the emptying out of meaning a glimpse of the materiality of the letter. In commenting on the first sentence de Man plays with mein and meinen as mine and mean and generates a sentence in which the cascade of "sinces," and sinces within sinces, produces its own stuttering repeti- tion, like a broken record:
"Ich kann nicht sagen was ich (nur) meine" then means "I cannot say what I make mine" or, since to think is to make mine, "I cannot say what I think," and, since to think is fully contained in and defined by the I, since Hegel's ego cogito defines itself as mere ego, what the sen- tence actually says is "I cannot say I"--a disturbing proposition in Hegel's own terms since the very possibility of thought depends on the possibility of saying "I. " (AI 98)
The other sentence, with its repetitions of ich and ich in mich, is already "astonishing" enough itself, as de Man says, in the sense of numbing the mind, turning it to stone (to play on a false etymology; the word really means, etymologically, "to strike with thunder"). The sentence shows the impossibility not only of the deictics "here," "now," "this," as when I say, "This sentence which I am here and now writing on my computer at 8:51 a. m. on November 4, 1997," or, in Hegel's example, this piece of paper on which I am now writing, but also of the deictic use of "I" to point to me myself alone as a unique I. These words are "shifters," placeholders. Instantly, as soon as they are uttered, the words assume the utmost generality and can be shifted to any I, any
here, now, and this. 12 However hard you try, you cannot say this I here and now or this keyboard, processor, and computer screen at this mo- ment that are prostheses of my body and by means of which I think. "I cannot say I. " "Aber was ich sage, Ich, ist eben jeder (but what I say, I, is precisely anyone). " De Man takes the otherness of "jeder" not to refer to another I, "the mirror image of the I," but to name "n'importe qui or even n'importe quoi" (AI 98); that is, anybody at all or even anything at all, just as the name Marion, in de Man's reading of the "purloined ribbon" episode in Rousseau's Julie, is ultimately just a random sound, not even a proper name: "Rousseau was making what- ever noise happened to come into his head; he was saying nothing at all, least of all someone's name" (AR 292).
As de Man says of Rousseau's excuse in Julie for what he had done to Marion, "When everything fails, one can always plead insanity" (AR 289). A certain madness, the madness of words, the reader can see, often infects de Man's own language. He mimes in what he says the materiality of the letter he is naming. At this point his own work becomes a performative utterance working to lead the reader to the edge of unintelligibility, this time by the route of the materiality of the letter, and once more in a way that is counterintuitive, since it is another materiality that is nonphenomenal, unable to be seen, like the "cristal invisible" of that Flemish windowpane on which Hugo's poem was scratched.
The back cover of de Man's Aesthetic Ideology speaks of the "ironic good humor that is unique to him. " I find de Man's irony, especially when it expresses itself in wordplay, much more threatening than this phrase implies, and so have many of de Man's readers or listeners. Such passages as I have been discussing, where the madness of words has crossed over into de Man's own language, are places that readers or auditors have found especially allergenic, that they have especially resisted. The audience of de Man's "Semiology and Rhetoric," for ex- ample, when the essay was presented as a sort of inaugural lecture after de Man took up his professorship at Yale, was more than a little scandalized or even offended by the elaborate pun de Man develops based on the Archie Bunker television show. This pun depends on the difference between lacing your shoes over or under. ("What's the dif- ference? " asks Archie Bunker. ) This leads to the punch line of calling Jacques Derrida an "archie Debunker" (AR 10). The audience did not find that wholly appropriate for such a solemn occasion. The complex double talk that de Man, in an exuberant reading, finds in Proust's
Paul de Man as Allergen 197
198 J. Hillis Miller
phrase "torrent d'activite? " (AR 64) has seemed to some readers just going too far. Raymond Geuss especially resisted what de Man says about "mein" and "meinen" in Hegel. De Man's "Reply to Raymond Geuss" patiently laces over and under, that is, explains what he meant and why he is right and Geuss wrong, guilty of "misplaced timidity" (AI 190), an unwillingness to face up to what is truly wild in Hegel's text.
The resistance to de Man, what I have called an allergic reaction to his writings, is not a resistance to theory in the etymological sense of the word theory, a resistance to a generalizable "clear-seeing," but rather a resistance to what in his work precisely cannot be seen clearly, the penumbra of the unknowable, the unintelligible, the nonphenomenal that is everywhere in his work. This is perhaps most threateningly pres- ent not in the radical incompatibility of the cognitive and performative dimensions of language, and not even in what Friedrich Schlegel called the madness and stupidity reached by irony as permanent parabasis, nor even in Kant's materiality of vision, but in the prosaic materiality of the letter. The latter is present at every moment, though for the most part it is invisible, suppressed, covered over, in all those words that sur- round us all the time and that generate the reassuring ideologies in terms of which we live our lives. What is most threatening, most aller- genic, most truly frightening about de Man's writings, is the way they force their readers to confront a darkness of unknowability that is not just out there somewhere, beyond the circle of light cast by the desk's reading lamp. That would be bad enough, but this darkness has woven itself into the light of reason itself and into the "instrument" by which it expresses itself, language. "No degree of knowledge can ever stop this madness, for it is the madness of words" (RR 122).
PAUL DE MAN'S AUTHORITY
Another double genitive there: the authority Paul de Man exerts and the authority in whose name he speaks. This essay began by identifying what is insolent or outrageous about de Man's writings, namely, his calm, laconic assertions that all the basic assumptions of literary stud- ies as a discipline, along with all the greatest authorities in that disci- pline, are often just plain wrong. Where does de Man get his authority to say such things? In the light of my investigation of his materialism I propose now in conclusion three braided answers to the question of what justifies de Man to say what he says. All these may be inferred from de Man's own writing.
First, he might be imagined as replying that what he says, allergenic
as it is, is not his own willful desire to cause trouble, but something that just happens, through reading. De Man's work is all reading of some text or other, primarily canonical texts that are among the most revered and cherished in our tradition. Therefore all these outrageous statements are not de Man speaking, but him speaking in indirect dis- course for what his authors say. It is Shelley, not de Man, who says that nothing is connected to anything else. Hegel or Kleist, not de Man, who repeats the same words or syllables until they become senseless. It is not I, Paul de Man, speaking, but I speaking in the name of, with the authority, of my authors. As Chaucer says, "My auctor wol I folwen if I konne. "13 In the "Reply to Raymond Geuss," de Man says,
The move from the theory of the sign to the theory of the subject has nothing to do with my being overconcerned with the Romantic tradi- tion, or narcissistic, or ("c'est la me^me chose") too influenced by the French. It has, in fact, nothing to do with me at all but corresponds to an inexorable and altogether Hegelian move of the text. (AI 189)
Or, second appeal to authority, what I, Paul de Man, say happens through the rigor of critical reading. This rigor is something that pro- duces the generalizations of theory, something that is wholly rational, logical, transmissible, the product of rigorous thinking that might have been done by anyone with de Man's intelligence and learning. Theory grows out of reading and is authorized by it, though it is in a different register and even though theory and reading, as "The Resistance to Theory" shows, are not symmetrical. Although "the resistance to theo- ry is in fact a resistance to reading," nevertheless "rhetorical readings, like the other kinds, still avoid and resist the reading they advocate. Nothing can overcome the resistance to theory since theory is itself this resistance" (RT 15, 19). In the "Reply to Raymond Geuss," de Man asserts that the commentator should accept the "canonical reading" up to the point where something is encountered in the text that makes it impossible to go on accepting the canonical interpretation. De Man's formulations are couched in the language of ethical obligation and in- evitability: "should," "could," and "necessity. " The necessity arises from the reader's encounter with the text. What happens in reading happens, and it imposes implacable obligations on the reader that exceed the pre- suppositions both of the canonical reading and of "theory":
The commentator should persist as long as possible in the canonical read- ing and should begin to swerve away from it only when he encounters
Paul de Man as Allergen 199
200 J. Hillis Miller
difficulties which the methodological and substantial assertions of the system are no longer able to master. Whether or not such a point has been reached should be left open as part of an ongoing critical investiga- tion. But it would be naive to believe that such an investigation could be avoided, even for the best of reasons. The necessity to revise the canon arises from resistances encountered in the text itself (extensively con- ceived) and not from preconceptions imported from elsewhere. (AI 186)
Third source of de Man's authority, deepest and most serious: the scandalous, counterintuitive things de Man says come into language through the encounter, at the limits of the most exigent theoretical rigor and obedient close reading, of the unintelligible. De Man takes the rational to the edge of irrationality, or identifies the unintelligible as that which has always already infected the pursuit of rational knowledge: "after Nietzsche (and, indeed, after any 'text'), we can no longer hope ever 'to know' in peace" (AR 126). Wherever de Man starts, whatever texts he reads, whatever vocabulary he uses leads ulti- mately beyond itself to its limits at the border of a dark unintelligibili- ty, what Friedrich Schlegel called "der Schein des Verkehrten und Verru? ckten oder des Einfa? ltigen und Dummen" ("the appearance of error and madness, or simplemindedness and stupidity"). 14 Three names de Man gives this unintelligibility are performative language, irony, and materiality. Kant may be taken as the paradigmatic model here. Kant's rigor of critical thinking led him to what undid his enter- prise of architectonic articulation, disarticulated it. The same thing can be said of de Man's writing, except that de Man's writing is throughout a long meditation on what happens when thinking encounters that momentary event when the unintelligible, error, madness, stupidity, undoes the rational enterprise of critical thinking, or turns out to have been undoing it all along.
De Man speaks in the name of, on the grounds of, these three quite incompatible but nevertheless inextricably intertwined justifications for the allergens that he generates in words. This authority is, however, no authority in the ordinary sense. It is an authority without authority, or the authority that undoes all grounds for speaking with authority. How can one speak intelligibly on the grounds of the unintelligible? At the limit, and indeed all along the way, de Man's writings are allergenic because they pass on to the reader an allergen, an otherness, with which they have been infected and that is quite other to the calm, im- placable, rational, maddeningly difficult to refute,15 rigor of de Man's
argumentation. Or rather, the latter turns out to be the same as the for- mer, reason to be other to itself.
NOTES
1. Paul de Man, Aesthetic Ideology, ed. Andrzej Warminski (Minneapolis: University of Minnesota Press, 1996), 131-32; hereafter AI.
2. Paul de Man, The Rhetoric of Romanticism (New York: Columbia University Press, 1984), 262; hereafter RR.
3. I use the word joyfully as an allusion to Nietzsche's "joyful wisdom" or "fro? hliche Wissenschaft. " Anyone who fails to see the exuberant or even comic joy in de Man's writings, anyone who sees him as a "gloomy existentialist," as one commentator calls him, simply lacks an ear. The ironic comedy sometimes surfaces openly, as when he says, apropos of Kant's assertion that the Dutch are all phleg- matic, "interested only in money and totally devoid of any feeling for beauty or sublimity whatsoever": "I have never felt more grateful for the hundred or so kilo- meters that separate Antwerp [de Man's home city] from Rotterdam" (AI 124-25). Another example is what he says as part of an assertion that the self-undoing of Kant's critical enterprise through "the rigor of his own discourse" was not felt as a subjective, affective shudder: "I don't think that Kant, when he wrote about the heavens and the sea there, that he was shuddering in mind. Any literalism there would not be called for. It is terrifying in a way which we don't know. What do we know about the nightmares of Immanuel Kant? I'm sure they were . . . very interesting . . . Ko? nigsberg there in the winter--I shudder to think (AI 134). This joy is no doubt one of the things that is held against de Man, as Derrida's exuberant hijinks--in format, for example--are held against him. Both make ironic jokes about deadly serious matters. There is no room for comedy or for joy either in phi- losophy and theory. They are solemn matters for which you should, if you are a man, always wear a shirt and tie.
4. Paul de Man, The Resistance to Theory (Minneapolis: University of Minne- sota Press, 1986), 3-4; hereafter RT; and RR 121-23.
5. De Man goes on to recognize that the final twist of irony in Greek or Hellenic comedy is that the smart guy is "always being set up by the person he thinks of as being the dumb guy, the alazon. In this case the alazon (and I recognize that this makes me the real alazon of this discourse) is American criticism of irony, and the smart guy is going to be German criticism of irony, which I of course under- stand" (AI 165). This seems to be a rare example of an overt admission by de Man that he is bound to be caught in the traps he sets for others, that what is sauce for the goose is sauce for the gander. In the rest of "The Concept of Irony," however, de Man allows precious little in the way of smart-guy attributes either to American criticism of irony, exemplified by Wayne Booth, presented as a dumb guy through and through, or to German and Danish criticism of irony either, with the exception of Friedrich Schlegel. Hegel, Kierkegaard, Benjamin, Szondi, and so on, are all as dumb as Booth, though in different ways. In the vibrating irony of the passage I have quoted from de Man, it is ironic for de Man to claim that he represents American criticism of irony, though of course he is not German either. In any case, for him to say he is "the real alazon of this discourse" is at the same time to say that
Paul de Man as Allergen 201
202 J. Hillis Miller
he is the real eiron, since the alazon always turns out to be the disguised eiron, the smartest smart guy, or the only smart guy around.
6. Paul de Man, Allegories of Reading (New Haven: Yale University Press, 1979), 78; hereafter AR.
7. The anonymous reader of this essay for the University of Minnesota Press strongly resisted this account of de Man's concept of historical events in their mate- riality. "Miller's idea of history, moreover," the reader said, "is of little merit and has, as far as I can tell, very little to do with de Man. " This is a good example of what I mean by an allergic reaction. My own idea of history is not expressed any- where here, only de Man's, although in the sentence beginning "History is wars, battles . . . " I am miming ironically what history is conventionally assumed to be. Can the reader have taken my irony straight? After a careful rereading of my essay, I claim that the citations from de Man I make support what I say about de Man's concept of history. It is de Man's concept, not mine, that scandalizes the reader, makes him (or her) sneeze and cough. I have, however, altered one phrase that ap- parently misled the reader into thinking I understand de Man to be saying that his- tory is caused by "intentional" uses of language and that might therefore mislead you, dear reader. As any careful reader of de Man knows, his theory of the perfor- mative "use" of language (as opposed to its mention) is detached from any con- scious intention in the user. Language works performatively, on its own, most often against the intentions or knowledge of the speaker or writer. As he says, in the con- clusion to "Promises (Social Contract)," "The error is not within the reader; lan- guage itself dissociates the cognition from the act. Die Sprache verspricht (sich)" (AR 277), which means "Language promises" and also "Language makes a slip of the tongue. " I have thought it worthwhile to refer directly to the comments of the Minnesota reader in order to try to forestall similar errors on the part of readers of the published essay.
8. Louis Althusser, "Ideology and Ideological State Apparatuses (Notes to- wards an Investigation," in Lenin and Philosophy and Other Essays, trans. Ben Brewster (New York: Monthly Review Press, 1972), 162. See page 159, where Althusser says, "ideology has no history," and goes on to remark: "As we know, this formulation appears in so many words in a passage from The German Ideology. "
9. In an equally important, though much less well known, definition of ideolo- gies near the beginning of "Phenomenality and Materiality in Kant" de Man assert- ed that ideologies are on the side of what Kant called "metaphysics," that is, in Kant's use of the term, precritical empirical knowledge of the world. Only critical analysis of ideologies will keep ideologies from becoming mere illusion and critical philosophy from becoming idealism cut off from the empirical world (AI 72). The anonymous reader for the University of Minnesota Press sternly challenged my understanding in this footnote of Kant's use of the term metaphysics. This is another allergic reaction, one that demonstrates just the point I am making about de Man. Surely Kant cannot have meant something so strange as this by "metaphysics"! At the risk of making this footnote tediously long for those who have read Kant and de Man's commentary on Kant, here is the relevant passage from Kant, followed by de Man's comment on it. I think my reader is mystified through having accepted received opinion about what Kant must be saying because everyone knows that is what he says. That received opinion is, precisely, a species of "ideology," even of
"aesthetic ideology. " Kant says: "A transcendental principle is one through which we represent a priori the universal condition under which alone things can become Objects of our cognition generally. A principle, on the other hand, is called meta- physical [Dagegen heisst ein Prinzip metaphysisch], where it represents a priori the condition under which alone Objects whose concept has to be given empirically [empirisch], may become further determined [bestimmet] a priori. Thus the prin- ciple of the cognition of bodies [der Erkenntnis der Ko? rper] as substances, and as changeable substances, is transcendental where the statement is that their change must have a cause [Ursache]: but it is metaphysical where it asserts that their change must have an external cause [eine a? ussere Ursache]. For in the first case bod- ies need only be thought through ontological predicates (pure concepts of under- standing [reine Verstandesbegriffe]), e. g. as substance, to enable the proposition to be cognized a priori; whereas, in the second case, the empirical concept of a body (as a movable thing in space) must be introduced to support the proposition [diesem Satze zum Grunde gelegt werden muss], although, once this is done, it may be seen [eingesehen] quite a priori that the latter predicate (movement only by means of an external cause) applies to body" (Immanuel Kant, Kritik der Urteilskraft, ed. Wilhelm Wieschedel [Frankfurt am Main: Suhrkamp, 1979], 90; The Critique of Judgement, trans. James Creed Meredith [Oxford: Oxford University Press, 1982], 20-21). De Man comments, in "Phenomenality and Materiality in Kant": "The condition of existence of bodies is called substance; to state that substance is the cause of the motion of bodies (as Kant does in the passage quoted) is to exam- ine critically the possibility of their existence. Metaphysical principles, on the other hand, take the existence of their object for granted as empirical fact. They contain knowledge of the world, but this knowledge is precritical. Transcendental prin- ciples contain no knowledge of the world or anything else, except for the knowl- edge that metaphysical principles that take them for their object are themselves in need of critical analysis, since they take for granted an objectivity that, for the tran- scendental principles, is not a priori available. Thus the objects of transcendental principles are always critical judgments that take metaphysical knowledge for their target. Transcendental philosophy is always the critical philosophy of metaphysics" (AI 71). De Man goes on to associate ideology with metaphysics as Kant defines it. The passage is an important gloss on de Man's definition, or, more properly, "call- ing," of ideology in "The Resistance to Theory," just cited. In the sentences that follow just after the ones already quoted from "Phenomenality and Materiality in Kant" de Man associates ideology with Kantian "metaphysics" and argues for an intricate interdependence of critical thought on ideology and of ideology, if it is to other than "mere error," on critical thought.
Almost any page of de Man's work, but especially the beginnings and endings of essays, contains rejections of well-established received ideas about literary study. These rejections can best be characterized as ironically and joyfully insolent or even contemptuous, as well as dis- mayingly rigorous and plausible. 3 Salient examples are the first two pages of "The Resistance to Theory" and the last three pages of "Shelley Disfigured. "4 De Man's essays have the structure he identifies in "The Concept of Irony" as "the traditional opposition between eiron and alazon, as they appear in Greek or Hellenic comedy, the smart guy and the dumb guy" (AI 165). De Man is of course the eiron, the smart guy, and all the previous experts on whatever topic or text he is discussing are the alazons, the dumb guys. 5 The received ideas he attacks, often fundamental assumptions of our profession, are charac- teristically called aberrant, deluded, or simply false. The reader can only hope or assume that "This does not, cannot, mean me! Surely I would not make such stupid mistakes. " De Man forestalls that defen- sive move, however, when he asserts, for example, in the "Kant and Schiller" essay in Aesthetic Ideology, that everyone, including himself,
however ironically, in a collective "we," is still bewitched by aesthetic ideology:
Before you either contest this [what he has been saying about Schiller's distortion of Kant], or before you not contest but agree with it and hold it against Schiller, or think that it is something we are now far beyond and that we would never in our enlightened days do--you would never make this naive confusion between the practical and the pragmatic on the one hand and the philosophical Kantian enterprise on the other-- before you decide that, don't decide too soon that you are beyond Schiller in any sense. I don't think any of us can lay this claim. Whatever writing we do, whatever way we have of talking about art, whatever way we have of teaching, whatever justification we give ourselves for teaching, whatever the standards are and the values by means of which we teach, they are more than ever and profoundly Schillerian. They come from Schiller, and not from Kant. (AI 142)
De Man goes on to make a warning that certainly applies to what has happened in his own case, in spite of the fact that he was protected by being a Sterling Professor at Yale, which is about as much security as you can get:
And if you ever try to do something in the other direction [in the direc- tion of Kant, that is, rather than Schiller] and you touch on it you'll see what will happen to you. Better be very sure, wherever you are, that your tenure is very well established, and that the institution for which you work has a very well-established reputation. Then you can take some risks without really taking many risks" (AI 142).
I have said that de Man's work is threatening to "us all" because al- most any page contains cheerfully taunting rejections, explicit or im- plicit, of "our" most basic ideological assumptions, the ones "we" most need to get on with our work, the ones the university most needs to get on with its work. His counterintuitive concept (it is not really a concept) of materiality is an example of this.
DE MAN'S MATERIALISM
The "'s" in this subhead is a double genitive, both objective and sub- jective. It names both de Man's theory of materiality and the way his own writings may show materiality at work or may be examples of materiality at work. De Man's materiality is one of the most difficult and obscure parts of his work.
Paul de Man as Allergen 185
186 J. Hillis Miller
De Man's use of the terms materiality and materialism poses several special problems, resistances to comprehension. First, one or the other word is most often introduced only briefly and elliptically. If the reader does not keep a sharp eye out for it, it appears in a given essay for an instant, for the blink of an eye, like a meteor, and then vanishes. Moreover, in these passages de Man seems to be saying exceedingly strange things, such as the assertion that materiality is not "phenome- nal. " Second, unlike "performative" and "irony" (terms not on every- one's lips and concepts that clearly need some explaining), we tend to think we already know what materiality is. It is the property possessed by these hard objects right in front of me now, impassive, impassible, resistant, not dependent on my perception for their continued exis- tence, like that stone Samuel Johnson kicked to refute Berkeley's ideal- ism: "I refute him thus [kicking the stone]. " Third, the term material- ism is extremely difficult to extricate from its associations with modern empirical science or with vulgar understandings of Marxism. Is not Marxism to be defined as "dialectical materialism"? De Man is sup- posed to be in one way or another a linguistic formalist, someone who believed, as all so-called deconstructionists are supposed to believe, that it is "all language," though the reader might remember that de Man began his higher education as a science, mathematics, and engi- neering student at the E? cole Polytechnique of the University of Brussels (1936). His professional interest in language came later. Nevertheless, for de Man to call himself a materialist, or for us to call him one, seems as absurd and counterintuitive as for de Man to call Kant and Hegel materialists or to find crucial materialist moments in their work, since everybody knows (without necessarily having read them) that they are "idealists. " Equally absurd would be to think one might find any kin- ship between de Man's thinking and Marxism, though the truth is that a deep kinship exists between de Man's work and Marx's thought in The German Ideology, as Andrzej Warminski has been demonstrating in his seminars. To show this it is necessary actually to go back and read Marx, as well as de Man, no easy tasks.
The term materiality or its cognates appears at crucial moments in de Man's work as early as a citation from Proust in "Reading (Proust)" in Allegories of Reading. What Proust calls the "symbols," in Giotto's Allegory of the Virtues and Vices at the Arena in Padua, meaning rep- resentations like the Charity that looks like a kitchen maid, are "some- thing real, actually experienced or materially handled. "6 That this pas- sage was important to de Man is indicated by the way he cites it again
at a crucial moment on the symbol in Hegel just at the end of one of his late essays, "Sign and Symbol in Hegel's Aesthetics. " This time de Man translates the phrases himself somewhat differently from the Moncrieff translation, and he cites the French original: "the symbol represented as real, as actually inflicted or materially handled [. . . (le symbole repre? sente? ) comme re? el, comme effectivement subi ou mate? riellement manie? ]" (AI 103). The terms material, materiality, and the like then appear with increasing frequency in de Man's later work. It is as though de Man had discovered in such words a way to "call" more ac- curately something he wanted performatively to name, perhaps even to invoke, that is, to "call forth": "The only word that comes to mind is that of a material vision . . . " ("Phenomenality and Materiality in Kant," in AI 82). What Michael Riffatere misses or evades in Hugo's "E? crit sur la vitre d'une fene^tre flamande" is just what the title indi- cates or names, namely, what de Man calls "the materiality of an in- scription" (RT 51). A climactic passage in Shelley's The Triumph of Life is said to stress "the literal and material aspects of language" (RR 113). "Anthropomorphism and Trope in the Lyric" ends, in a phrase I have already cited, with an appeal to "the materiality of actual history" (RR 262). A cascade of such terms punctuates the essays in Aesthetic Ideology, not only in "Phenomenality and Materiality in Kant" and in "Kant's Materialism," where "a materialism that Kant's posterity has not yet begun to face up to" (AI 89) is the focus of the argument, but also in "Sign and Symbol in Hegel's Aesthetics," where we read that "The idea, in other words, makes its sensory appearance, in Hegel, as the material inscription of names" and also in the way Hegel's "theory of the sign manifests itself materially" (AI 102, 103), and in "Kant and Schiller," where we read of the irreversible progression "from states of cognition, to something which is no longer a cognition but which is to some extent an occurrence, which has the materiality of something that actually happens, that actually occurs" and of "the materiality of the inscribed signifier in Kant" (AI 132, 134).
The reader will have seen that the term materiality and its cognates occur in three related, ultimately more or less identical, registers in de Man: the materiality of history, the materiality of inscription, and the materiality of what the eye sees prior to perception and cognition. In all three of these registers, as I shall try to show, materiality is associat- ed with notions of performative power and with what seems materiali- ty's opposite, formalism. In all three modes of materialism, the ultimate paradox, allergenic idea, or unintelligibility is the claim or insinuation
Paul de Man as Allergen 187
188 J. Hillis Miller
that materiality is not phenomenal, not open to the senses. Just what in the world could that mean?
The phrase "materiality of history" seems the easiest to understand and accept as commonsensical. Of course history is material. It means what really happened, especially as a result of human intervention (though we speak, for example, of the history of the mollusks, or of geo- logical history). History is wars, battles, the building of the pyramids, the invention of the steam engine, migrations of peoples, legislative de- cisions, diplomatic negotiations, the clearing of forests, global warm- ing, that sort of thing. De Man's materiality of history, however, is not quite like that. For him the materiality of history, properly speaking, is the result of acts of power that are punctual and momentary, since they are atemporal, noncognitive and noncognizable performative utter- ances. History is caused by language or other signs that make some- thing materially happen, and such happenings do not happen all that often. The most radical, and allergenic, counterintuitive, scandalous formulation of this is in "Kant and Schiller. " There de Man asserts that Kant's Critique of Judgment was an irreversible historical event brought about by the shift from cognitive to efficaciously performa- tive discourse in Kant's own words, whereas Schiller's ideological mis- reading of Kant and its long progeny in the nineteenth and twentieth centuries were nonevents, certainly not irreversible material events. In "Phenomenality and Materiality in Kant" de Man speaks of the crucial shift to a "formal materialism" in Kant's Critique of Judgment as "a shift from trope to performance" that is "a deep, perhaps fatal, break or discontinuity" (AI 83, 89, 79). This is the place, as he puts it in "Kant and Schiller," at which Kant "found himself by the rigor of his own discourse [the project of aesthetics as articulation of pure reason and practical reason or ethics] to break down under the power of his own critical epistemological discourse" (AI 134). This was an event, strictly speaking an irreversible historical event, "to some extent an oc- currence, which has the materiality of something that actually hap- pens, that actually occurs. And there, the thought of material occur- rence, something that occurs materially, that leaves a trace on the world, that does something to the world as such--that notion of oc- currence is not opposed in any sense to the notion of writing" (AI 132). Since the event of Kant's materialism is punctual and instanta- neous, it is in a curious sense not within time, though it has a perma- nent and irreversible effect on what we usually (mistakenly) think of as the temporality of history:
history is not thought of as a progression or a regression, but is thought of as an event, as an occurrence. There is history from the moment that words such as 'power' and 'battle' and so on emerge on the scene. At that moment things happen, there is occurrence, there is event. History is therefore not a temporal notion, it has nothing to do with temporality [there's allergenic assertion for you! ], but it is the emergence of a lan- guage of power out of a language of cognition. (AI 133)
I do not think de Man meant that the words power and battle are in themselves always historical events in the sense de Man is defining such events, but that he means the uses of such words in effective per- formative utterances are historical events. As opposed to the moment of Kant's self-undoing materialism in the third Critique, Schiller's recu- peration of Kant within aesthetic ideology and its long progeny, the procedures of which are identified in the main body of "Kant and Schiller," did not happen, were not historical events:
One could say, for example, that in the reception of Kant, in the way Kant has been read, since the third Critique--and that was an occur- rence, something happened there, something occurred [de Man's stut- tering iterations here mime the punctualities of historical events; the reader will remember that this is the transcript of an oral presentation that was not written down as such]--that in the whole reception of Kant from then until now, nothing has happened, only regression, noth- ing has happened at all. Which is another way of saying there is no his- tory . . . that reception is not historical. . . . The event, the occurrence, is resisted by reinscribing it in the cognition of tropes, and that is itself a tropological, cognitive, and not a historical move. (AI 134)7
These sternly recalcitrant statements may be more understandable and perhaps even more acceptable if we remember that Althusser, and de Man in his own way, following Marx, define ideology as having no history, as being outside history, as having no purchase on his- tory, since ideology is precisely an illusory misunderstanding of the "real conditions of existence," as Althusser put it in "Ideology and Ideological State Apparatuses,"8 or, as de Man puts this in "The Resis- tance to Theory": "What we call ideology is precisely the confusion of linguistic with natural reality, of reference with phenomenalism" (RT 11). 9 The reception of Kant by Schiller and his followers, including you and me as inheritors of aesthetic ideology, is ideological, therefore not historical.
Paul de Man as Allergen 189
190 J. Hillis Miller
We are (I am) now in a position to answer the puzzling assertions de Man makes in "The Concept of Irony. " "Irony," he says, "also very clearly has a performative function. Irony consoles and it promises and it excuses" (AI 165). What could de Man mean by saying that irony is performatively efficacious, that it promises, consoles, or excuses? If we take seriously de Man's claim later in the essay that irony is a perma- nent parabasis that radically suspends meaning by the incursion of chaos, madness, and stupidity (Friedrich Schlegel's terms) into lan- guage, then it would seem radically counterintuitive to say that irony has a successful performative function. A statement at the end of the essay is equally baffling: "Irony and history seem to be curiously linked to one another" (AI 184). If irony is permanent parabasis it would seem to have little to do with history, but to be rather the withdrawal from effective historical action. The analogy between the noncognitive aspect of irony and the noncognitive aspect of performative utterances gives the clue. Irony is perhaps the most radical example of the rupture between cognitive and performative discourses. Insofar as an utterance is performative, it is unknowable. Irony suspends cognition. It is just because irony is error, madness, and stupidity that it can be performa- tively felicitous. Promises, excuses, consolations can be performed by irony, or can be especially done by ironic utterance, just because irony is the radical suspension of cognition. Another way to put this is to say that even the most solemn performative utterances are contaminated by being possibly ironic. Jacques Derrida includes irony along with lit- erature among the parasitical presences that are possibly incorporated within any performative as a result of its intrinsic iterability.
What I have just said will also indicate the surprising and "curious" connection of irony with history. Since the materiality of history as event is generated by acts of linguistic power, that is, performative speech acts, though by no means necessarily intentional ones, irony as a form of such power or as an ingredient of any such act of power, against all our instinctive assumptions, can be said not only to prom- ise, console, and excuse, but also to generate the events that make up the materiality of history. Just as, for Derrida, the possibility of felici- tous speech acts depends on the possibility that they may be "litera- ture," so for de Man the efficacy of performative utterances, including those that generate history, depends on the possibility that they may be ironical. They may be. You cannot tell for sure.
If speech acts generating history are, strangely enough, one form of materiality or are the place where language touches materiality, leaves
a mark on it, materially handles it, the materiality of what the eye sees appears more obvious but turns out to be more difficult to grasp. Of course, we say, what the eye sees is material. That received opinion or doxa turns out, however, once again not to be quite what de Man means. What he does mean is the central argument of the two essays on Kant, "Phenomenality and Materiality in Kant" and "Kant's Materialism. " For received opinion, what we take for granted, phe- nomenality and materiality are the same thing or are two aspects of the same thing. Because something is material it is phenomenal, open to the senses. For de Man, following Kant, phenomenality and materiali- ty are not conjoined but opposed. How can this be? De Man sees in Kant's theory of the dynamic sublime two radically contradictory no- tions. On the one hand, the sublime is the moment when the imagina- tion triumphs over fear and puts all the elements of the sublime scene together, articulates them in a grand aesthetic synthesis, as tropes ar- ticulate, or as the body's limbs are articulated: "The imagination over- comes suffering, becomes apathetic, and sheds the pain of natural shock. It reconciles pleasure with pain and in so doing it articulates, as mediator, the movement of the affects with the legal, codified, formal- ized, and stable order of reason" (AI 86). In so doing, the imagination of the sublime or the sublime itself accomplishes the goal of the third Critique, which was to find a "bridge" between the first and second Critiques, between pure reason and the practical reason of moral obli- gation and choice. On the other hand, Kant's analysis of the dynamic sublime contains a moment that radically disrupts, interrupts, and sus- pends this happy articulation. Kant reaches this moment through the very rigor of his critical thinking. He proposes that the paradigmatic example of the dynamic sublime is when the overarching vault of the sky and the outstretched mirror of the sea are seen just as the eye sees them, or as the poets see them, without thought for their meaning. Seeing them as meaningful would occur, for example, when we view the sea as a reservoir of edible fish, or the sky as a producer of life-giving rain. De Man quotes section 28 of Kant's The Critique of Judgment: "we must regard it [the starry heaven], just as we see it [wie man ihn sieht], as a distant, all-embracing vault [ein weites Gewo? lbe]. . . . To find the ocean nevertheless sublime we must regard it as poets do [wie die Dichter es tun], merely by what the eye reveals [was der Augeschein zeigt]" (AI 80). De Man goes on to argue that this way of seeing is radi- cally nonphenomenal. It does not involve the mind that in its activity of perception would make sense of what is seen. It just sees what it
Paul de Man as Allergen 191
192 J. Hillis Miller
sees, in an activity of the eye operating by itself, enclosed in itself, wholly detached, disarticulated, from thinking and interpreting: "No mind is involved in the Kantian vision of ocean and heaven. . . . That is how things are to the eye, in the redundancy of their appearance to the eye and not to the mind, as in the redundant word Augenschein, . . . in which the eye, tautologically, is named twice, as eye itself and as what appears to the eye" (AI 82). De Man's name for this way of seeing is "material vision": "The only word that comes to mind is that of a material vision" (AI 82), which is another way of saying, in a paradig- matic performative speech act, "I call this 'material vision. '" The word material then appears in a cascade of phrases in the subsequent pages: "the vision is purely material"; "what we call the material aspect"; "a materialism that, in the tradition of the reception of the third Critique, is seldom or never perceived"; "If the architectonic then appears, very near the end of the analytics of the aesthetic, at the conclusion of the section on the sublime, as the material disarticulation not only of na- ture but of the body [traditional examples of the beautiful or the sub- lime], then this moment marks the undoing of the aesthetic as a valid category. The critical power of a transcendental philosophy undoes the very project of such a philosophy leaving us, certainly not with an ideology--for transcendental and ideological (metaphysical) principles are part of the same system--but with a materialism that Kant's pos- terity has not yet begun to face up to" (AI 83, 88, 89).
How could we "face up to" something that we can see but not face up to in the sense of clearly confronting it and making it intelligible to ourselves? The idea of a way of seeing that is performed by the eye alone, wholly dissociated from the mind, is, strictly speaking, unintelli- gible, since any sense we give to this Augenschein is an illicit, ideologi- cal imposition: "To the extent that any mind, that any judgment, inter- venes, it is in error" (AI 82). That is what I mean by saying that de Man's materiality is nonphenomenal, since phenomenality always in- volves, instantly, making sense or trying to make sense of what we see. This "material vision" would be pure seeing prior to any seeing as the sort of understanding that we name when we say, "I see it all now. " It would be a pre-seeing seeing, that is, something unthinkable, unknow- able, unintelligible, a tautological eye eyeing: "Realism postulates a phenomenalism of experience which is here being denied or ignored. Kant's looking at the world just as one sees it ('wie man ihn sieht') is an absolute, radical formalism that entertains no notion of reference or semiosis" (AI 128).
The idea of a materiality that would not be phenomenal does not make sense. Nevertheless, that is just what de Man affirms, most overtly and in so many words at the end of the essay on Riffaterre, "Hypogram and Inscription. " There he speaks of "the materiality (as distinct from the phenomenality) that is thus revealed [when we remember that Hugo's poem was supposed to have been written on a window pane], the unseen 'cristal' whose existence thus becomes a certain there and a certain then which can become a here and a now in the reading 'now' taking place" (RT 51). The paradox is that the window glass, figure here for the materiality of inscription, is not what the eye sees but what the eye sees through. In the Kant essays, as in "Hypogram and In- scription," the rigor of de Man's own critical thinking brings him re- peatedly, by different routes, across the border of the intelligible and into the realm of the allergenic, in this case the recognition of a materi- alism in Kant that has seldom or never been recognized in the whole distinguished tradition of Kant scholarship and so is anathema to it, just as de Man's reading of somewhat similar material moments in Hegel was anathema to the distinguished Hegel specialist Raymond Geuss. 10
The final version of materiality in de Man is the "prosaic materiali- ty of the letter" (AI 90). Just what does de Man mean by that? No one doubts that writing (and speaking too) have a material base, marks on paper or modulated waves in the air. This materiality is the benign base of the meaning, permanence, and transmissibility of language. No problem. De Man of course does not mean anything so in agreement with common sense and received opinion. When de Man calls Kant's sublime Augenschein of sky and sea a "material vision" he goes on to raise a further question that is not answered until the end of the essay: "how this materiality is then to be understood in linguistic terms is not, as yet, clearly intelligible" (AI 82). The answer is the materiality of the letter, but just what does that mean? The essay ends with an ex- planation that if not clearly intelligible, at least indicates why these "linguistic terms" must be unintelligible. The reader is given intelli- gence of unintelligibility, new news of the unknowable.
The prosaic materiality of the letter, linguistic "equivalent" of a ma- terialism of vision, has two main features. One is a disarticulation of language equaling the disarticulations of nature and the human body de Man has found in Kant's dynamic sublime: "To the dismemberment of the body corresponds a dismemberment of language, as meaning- producing tropes are replaced by the fragmentation of sentences and
Paul de Man as Allergen 193
194 J. Hillis Miller
propositions into discrete words, or the fragmentation of words into syllables or finally letters" (AI 89). Strictly speaking, as linguists, not to speak of language philosophers like Wittgenstein, have shown, words do not have meaning by themselves. They have meaning only when they are used, incorporated into sentences. To detach them from their sentences and leave them hanging there in the air or on the page, surrounded by blank paper, is the first stage in a progressive disarticu- lation of meaning that goes then to syllables and finally to letters. It is extremely difficult to see words, syllables, or letters, for example on a printed page, in this way, just as it is extremely hard to see as the eye sees. One has to be a poet, as Kant says, to do it. The mind instantly in- terprets what the eye sees, "perceives it," and gives meaning to it, just as the mind projects meaning into those mute letters on the page. It is almost impossible to see letters as just the material marks they are. Even words in a language we do not know are seen as language and not as sheer materiality. We tend to see random marks on a rock as possibly writing in an unknown language.
The other feature of the materiality of the letter stressed by de Man makes that materiality more likely to be glimpsed, in the wink of the eye, before the mind starts "reading. " This is repetition of words and word parts that calls attention to the absurd and unmotivated echoes among them at the level of syllable and letter: puns, rhymes, allitera- tions, assonances, and so on, that is, precisely those linguistic features poets especially use, "the play of the letter and of the syllable, the way of saying . . . as opposed to what is being said" (AI 89). The "persua- siveness" of the passage in Kant about the recovery of the imagination's tranquillity through material vision depends, de Man says, "on the proximity between the German words for surprise and admiration, Verwunderung and Bewunderung" (AI 89).
The reader, de Man contin- ues, is led to assent to the incompatibility or aporia between the imagi- nation's failure and its success by "a constant, and finally bewildering alternation of the two terms, Angemessen(heit) and Unangemessen(heit), to the point where one can no longer tell them apart" (AI 90). One ad- ditional example of this in de Man's essays is the cascade of words in "fall" that he finds in a passage by Kleist: Fall, Beifall, Su? ndenfall, Ru? ckfall, Einfall, Zuru? ckfall, Fa? lle: "As we know from another narra- tive text of Kleist ["On the Gradual Formation of Thoughts while Speaking"], the memorable tropes that have the most success (Beifall) occur as mere random improvisation (Einfall) at the moment when the author has completely relinquished control over his meaning and has
relapsed (Zuru? ckfall) into the extreme formalization, the mechanical predictability of grammatical declensions (Fa? lle)" (RR 290). By the time the reader gets to the end of this the root "fall" is fast becoming a mere surd, a sound emptied of meaning: "fall, fall, fall, fall. " The read- er will see that "formalism" of "formalization" names for de Man not the beautiful aesthetic formalization of the artwork, but a principle of mechanical senselessness in language that he associates with the arbi- trariness of grammar, of declensions, Fa? lle. De Man goes on to make a pun of his own. Since Falle also means trap in German, he can say that everyone falls into "the trap of an aesthetic education which inevitably confuses dismemberment of language by the power of the letter with the gracefulness of a dance. " That trap, however, is not a benign aes- theticizing of the random formalizations of language in grammar and paronomasia such as poets are known to play with. It is a mortal dan- ger, a pericolo de morte, according to the last words of the last essay in The Rhetoric of Romanticism, "the ultimate trap, as unavoidable as it is deadly" (RR 290). The reader will note that this aspect of the materi- ality of the letter tends to disappear in translation. It depends on the unique idiom, idiolect, or even "idiocy," in the etymological sense, of a certain language. Ultimately, this repetition of words and bits of words empties language of meaning and makes it mere unintelligible sound, as when the poet Tennyson, as a child, used to repeat his own name over and over, "Alfred, Alfred, Alfred," until it ceased to mean anything at all and he melted into a kind of oceanic trance. Try it with your own name, as I do here with mine: "Hillis, Hillis, Hillis, Hillis. "
De Man's formulation of this in one notable place is more prosaic. As he shows, Hegel's theory of memory as Geda? chtnis, in opposition to Erinnerung, is that it memorizes by emptying words of meaning and repeating them by rote, as pure arbitrary signs that might be in a foreign language or in no language at all:
"It is well known," says Hegel, "that one knows a text by heart [or by rote] only when one no longer associates any meaning with the words; in reciting what one thus knows by heart one necessarily drops all ac- centuation. " [I suppose Hegel means that one repeats the words mind- lessly, like a schoolchild or a robot--JHM. ] . . . The idea, in other words, makes its sensory appearance, in Hegel, as the material inscrip- tion of names. (AI 101-2)
Speaking in "Hegel on the Sublime" of Hegel's "Gesetz der A? usserlichkeit (law of exteriority)," de Man says, "Like a stutter, or a broken record,
Paul de Man as Allergen 195
196 J. Hillis Miller
it makes what it keeps repeating worthless and meaningless" (AI 116). This had already been exemplified in a truly vertiginous couple of paragraphs in "Sign and Symbol in Hegel's Aesthetics. " There de Man takes two at first innocent-enough-looking, but in fact "quite astonish- ing," sentences in Hegel's Encyclopedia: "Since language states only what is general, I cannot say what is only my opinion [so kann ich nicht sagen was ich nur meine]," and "When I say 'I,' I mean myself as this I to the exclusion of all others; but what I say, I, is precisely any- one; any I, as that which excludes all others from itself [ebenso, wenn ich sage: 'Ich,' meine ich mich als diesen alle anderen Ausschliessenden; aber was ich sage, Ich, ist eben jeder]" (AI 97, 98). The sentences themselves are bad enough in English, though worse in German (e. g. , wenn Ich sage, Ich, meine ich mich"), but by the time de Man gets through with these sentences the reader is dizzied by the repetitions, like Tennyson repeating his own first name, or as if he had been caught in a revolving door. 11 Through this dizziness the reader reaches in the emptying out of meaning a glimpse of the materiality of the letter. In commenting on the first sentence de Man plays with mein and meinen as mine and mean and generates a sentence in which the cascade of "sinces," and sinces within sinces, produces its own stuttering repeti- tion, like a broken record:
"Ich kann nicht sagen was ich (nur) meine" then means "I cannot say what I make mine" or, since to think is to make mine, "I cannot say what I think," and, since to think is fully contained in and defined by the I, since Hegel's ego cogito defines itself as mere ego, what the sen- tence actually says is "I cannot say I"--a disturbing proposition in Hegel's own terms since the very possibility of thought depends on the possibility of saying "I. " (AI 98)
The other sentence, with its repetitions of ich and ich in mich, is already "astonishing" enough itself, as de Man says, in the sense of numbing the mind, turning it to stone (to play on a false etymology; the word really means, etymologically, "to strike with thunder"). The sentence shows the impossibility not only of the deictics "here," "now," "this," as when I say, "This sentence which I am here and now writing on my computer at 8:51 a. m. on November 4, 1997," or, in Hegel's example, this piece of paper on which I am now writing, but also of the deictic use of "I" to point to me myself alone as a unique I. These words are "shifters," placeholders. Instantly, as soon as they are uttered, the words assume the utmost generality and can be shifted to any I, any
here, now, and this. 12 However hard you try, you cannot say this I here and now or this keyboard, processor, and computer screen at this mo- ment that are prostheses of my body and by means of which I think. "I cannot say I. " "Aber was ich sage, Ich, ist eben jeder (but what I say, I, is precisely anyone). " De Man takes the otherness of "jeder" not to refer to another I, "the mirror image of the I," but to name "n'importe qui or even n'importe quoi" (AI 98); that is, anybody at all or even anything at all, just as the name Marion, in de Man's reading of the "purloined ribbon" episode in Rousseau's Julie, is ultimately just a random sound, not even a proper name: "Rousseau was making what- ever noise happened to come into his head; he was saying nothing at all, least of all someone's name" (AR 292).
As de Man says of Rousseau's excuse in Julie for what he had done to Marion, "When everything fails, one can always plead insanity" (AR 289). A certain madness, the madness of words, the reader can see, often infects de Man's own language. He mimes in what he says the materiality of the letter he is naming. At this point his own work becomes a performative utterance working to lead the reader to the edge of unintelligibility, this time by the route of the materiality of the letter, and once more in a way that is counterintuitive, since it is another materiality that is nonphenomenal, unable to be seen, like the "cristal invisible" of that Flemish windowpane on which Hugo's poem was scratched.
The back cover of de Man's Aesthetic Ideology speaks of the "ironic good humor that is unique to him. " I find de Man's irony, especially when it expresses itself in wordplay, much more threatening than this phrase implies, and so have many of de Man's readers or listeners. Such passages as I have been discussing, where the madness of words has crossed over into de Man's own language, are places that readers or auditors have found especially allergenic, that they have especially resisted. The audience of de Man's "Semiology and Rhetoric," for ex- ample, when the essay was presented as a sort of inaugural lecture after de Man took up his professorship at Yale, was more than a little scandalized or even offended by the elaborate pun de Man develops based on the Archie Bunker television show. This pun depends on the difference between lacing your shoes over or under. ("What's the dif- ference? " asks Archie Bunker. ) This leads to the punch line of calling Jacques Derrida an "archie Debunker" (AR 10). The audience did not find that wholly appropriate for such a solemn occasion. The complex double talk that de Man, in an exuberant reading, finds in Proust's
Paul de Man as Allergen 197
198 J. Hillis Miller
phrase "torrent d'activite? " (AR 64) has seemed to some readers just going too far. Raymond Geuss especially resisted what de Man says about "mein" and "meinen" in Hegel. De Man's "Reply to Raymond Geuss" patiently laces over and under, that is, explains what he meant and why he is right and Geuss wrong, guilty of "misplaced timidity" (AI 190), an unwillingness to face up to what is truly wild in Hegel's text.
The resistance to de Man, what I have called an allergic reaction to his writings, is not a resistance to theory in the etymological sense of the word theory, a resistance to a generalizable "clear-seeing," but rather a resistance to what in his work precisely cannot be seen clearly, the penumbra of the unknowable, the unintelligible, the nonphenomenal that is everywhere in his work. This is perhaps most threateningly pres- ent not in the radical incompatibility of the cognitive and performative dimensions of language, and not even in what Friedrich Schlegel called the madness and stupidity reached by irony as permanent parabasis, nor even in Kant's materiality of vision, but in the prosaic materiality of the letter. The latter is present at every moment, though for the most part it is invisible, suppressed, covered over, in all those words that sur- round us all the time and that generate the reassuring ideologies in terms of which we live our lives. What is most threatening, most aller- genic, most truly frightening about de Man's writings, is the way they force their readers to confront a darkness of unknowability that is not just out there somewhere, beyond the circle of light cast by the desk's reading lamp. That would be bad enough, but this darkness has woven itself into the light of reason itself and into the "instrument" by which it expresses itself, language. "No degree of knowledge can ever stop this madness, for it is the madness of words" (RR 122).
PAUL DE MAN'S AUTHORITY
Another double genitive there: the authority Paul de Man exerts and the authority in whose name he speaks. This essay began by identifying what is insolent or outrageous about de Man's writings, namely, his calm, laconic assertions that all the basic assumptions of literary stud- ies as a discipline, along with all the greatest authorities in that disci- pline, are often just plain wrong. Where does de Man get his authority to say such things? In the light of my investigation of his materialism I propose now in conclusion three braided answers to the question of what justifies de Man to say what he says. All these may be inferred from de Man's own writing.
First, he might be imagined as replying that what he says, allergenic
as it is, is not his own willful desire to cause trouble, but something that just happens, through reading. De Man's work is all reading of some text or other, primarily canonical texts that are among the most revered and cherished in our tradition. Therefore all these outrageous statements are not de Man speaking, but him speaking in indirect dis- course for what his authors say. It is Shelley, not de Man, who says that nothing is connected to anything else. Hegel or Kleist, not de Man, who repeats the same words or syllables until they become senseless. It is not I, Paul de Man, speaking, but I speaking in the name of, with the authority, of my authors. As Chaucer says, "My auctor wol I folwen if I konne. "13 In the "Reply to Raymond Geuss," de Man says,
The move from the theory of the sign to the theory of the subject has nothing to do with my being overconcerned with the Romantic tradi- tion, or narcissistic, or ("c'est la me^me chose") too influenced by the French. It has, in fact, nothing to do with me at all but corresponds to an inexorable and altogether Hegelian move of the text. (AI 189)
Or, second appeal to authority, what I, Paul de Man, say happens through the rigor of critical reading. This rigor is something that pro- duces the generalizations of theory, something that is wholly rational, logical, transmissible, the product of rigorous thinking that might have been done by anyone with de Man's intelligence and learning. Theory grows out of reading and is authorized by it, though it is in a different register and even though theory and reading, as "The Resistance to Theory" shows, are not symmetrical. Although "the resistance to theo- ry is in fact a resistance to reading," nevertheless "rhetorical readings, like the other kinds, still avoid and resist the reading they advocate. Nothing can overcome the resistance to theory since theory is itself this resistance" (RT 15, 19). In the "Reply to Raymond Geuss," de Man asserts that the commentator should accept the "canonical reading" up to the point where something is encountered in the text that makes it impossible to go on accepting the canonical interpretation. De Man's formulations are couched in the language of ethical obligation and in- evitability: "should," "could," and "necessity. " The necessity arises from the reader's encounter with the text. What happens in reading happens, and it imposes implacable obligations on the reader that exceed the pre- suppositions both of the canonical reading and of "theory":
The commentator should persist as long as possible in the canonical read- ing and should begin to swerve away from it only when he encounters
Paul de Man as Allergen 199
200 J. Hillis Miller
difficulties which the methodological and substantial assertions of the system are no longer able to master. Whether or not such a point has been reached should be left open as part of an ongoing critical investiga- tion. But it would be naive to believe that such an investigation could be avoided, even for the best of reasons. The necessity to revise the canon arises from resistances encountered in the text itself (extensively con- ceived) and not from preconceptions imported from elsewhere. (AI 186)
Third source of de Man's authority, deepest and most serious: the scandalous, counterintuitive things de Man says come into language through the encounter, at the limits of the most exigent theoretical rigor and obedient close reading, of the unintelligible. De Man takes the rational to the edge of irrationality, or identifies the unintelligible as that which has always already infected the pursuit of rational knowledge: "after Nietzsche (and, indeed, after any 'text'), we can no longer hope ever 'to know' in peace" (AR 126). Wherever de Man starts, whatever texts he reads, whatever vocabulary he uses leads ulti- mately beyond itself to its limits at the border of a dark unintelligibili- ty, what Friedrich Schlegel called "der Schein des Verkehrten und Verru? ckten oder des Einfa? ltigen und Dummen" ("the appearance of error and madness, or simplemindedness and stupidity"). 14 Three names de Man gives this unintelligibility are performative language, irony, and materiality. Kant may be taken as the paradigmatic model here. Kant's rigor of critical thinking led him to what undid his enter- prise of architectonic articulation, disarticulated it. The same thing can be said of de Man's writing, except that de Man's writing is throughout a long meditation on what happens when thinking encounters that momentary event when the unintelligible, error, madness, stupidity, undoes the rational enterprise of critical thinking, or turns out to have been undoing it all along.
De Man speaks in the name of, on the grounds of, these three quite incompatible but nevertheless inextricably intertwined justifications for the allergens that he generates in words. This authority is, however, no authority in the ordinary sense. It is an authority without authority, or the authority that undoes all grounds for speaking with authority. How can one speak intelligibly on the grounds of the unintelligible? At the limit, and indeed all along the way, de Man's writings are allergenic because they pass on to the reader an allergen, an otherness, with which they have been infected and that is quite other to the calm, im- placable, rational, maddeningly difficult to refute,15 rigor of de Man's
argumentation. Or rather, the latter turns out to be the same as the for- mer, reason to be other to itself.
NOTES
1. Paul de Man, Aesthetic Ideology, ed. Andrzej Warminski (Minneapolis: University of Minnesota Press, 1996), 131-32; hereafter AI.
2. Paul de Man, The Rhetoric of Romanticism (New York: Columbia University Press, 1984), 262; hereafter RR.
3. I use the word joyfully as an allusion to Nietzsche's "joyful wisdom" or "fro? hliche Wissenschaft. " Anyone who fails to see the exuberant or even comic joy in de Man's writings, anyone who sees him as a "gloomy existentialist," as one commentator calls him, simply lacks an ear. The ironic comedy sometimes surfaces openly, as when he says, apropos of Kant's assertion that the Dutch are all phleg- matic, "interested only in money and totally devoid of any feeling for beauty or sublimity whatsoever": "I have never felt more grateful for the hundred or so kilo- meters that separate Antwerp [de Man's home city] from Rotterdam" (AI 124-25). Another example is what he says as part of an assertion that the self-undoing of Kant's critical enterprise through "the rigor of his own discourse" was not felt as a subjective, affective shudder: "I don't think that Kant, when he wrote about the heavens and the sea there, that he was shuddering in mind. Any literalism there would not be called for. It is terrifying in a way which we don't know. What do we know about the nightmares of Immanuel Kant? I'm sure they were . . . very interesting . . . Ko? nigsberg there in the winter--I shudder to think (AI 134). This joy is no doubt one of the things that is held against de Man, as Derrida's exuberant hijinks--in format, for example--are held against him. Both make ironic jokes about deadly serious matters. There is no room for comedy or for joy either in phi- losophy and theory. They are solemn matters for which you should, if you are a man, always wear a shirt and tie.
4. Paul de Man, The Resistance to Theory (Minneapolis: University of Minne- sota Press, 1986), 3-4; hereafter RT; and RR 121-23.
5. De Man goes on to recognize that the final twist of irony in Greek or Hellenic comedy is that the smart guy is "always being set up by the person he thinks of as being the dumb guy, the alazon. In this case the alazon (and I recognize that this makes me the real alazon of this discourse) is American criticism of irony, and the smart guy is going to be German criticism of irony, which I of course under- stand" (AI 165). This seems to be a rare example of an overt admission by de Man that he is bound to be caught in the traps he sets for others, that what is sauce for the goose is sauce for the gander. In the rest of "The Concept of Irony," however, de Man allows precious little in the way of smart-guy attributes either to American criticism of irony, exemplified by Wayne Booth, presented as a dumb guy through and through, or to German and Danish criticism of irony either, with the exception of Friedrich Schlegel. Hegel, Kierkegaard, Benjamin, Szondi, and so on, are all as dumb as Booth, though in different ways. In the vibrating irony of the passage I have quoted from de Man, it is ironic for de Man to claim that he represents American criticism of irony, though of course he is not German either. In any case, for him to say he is "the real alazon of this discourse" is at the same time to say that
Paul de Man as Allergen 201
202 J. Hillis Miller
he is the real eiron, since the alazon always turns out to be the disguised eiron, the smartest smart guy, or the only smart guy around.
6. Paul de Man, Allegories of Reading (New Haven: Yale University Press, 1979), 78; hereafter AR.
7. The anonymous reader of this essay for the University of Minnesota Press strongly resisted this account of de Man's concept of historical events in their mate- riality. "Miller's idea of history, moreover," the reader said, "is of little merit and has, as far as I can tell, very little to do with de Man. " This is a good example of what I mean by an allergic reaction. My own idea of history is not expressed any- where here, only de Man's, although in the sentence beginning "History is wars, battles . . . " I am miming ironically what history is conventionally assumed to be. Can the reader have taken my irony straight? After a careful rereading of my essay, I claim that the citations from de Man I make support what I say about de Man's concept of history. It is de Man's concept, not mine, that scandalizes the reader, makes him (or her) sneeze and cough. I have, however, altered one phrase that ap- parently misled the reader into thinking I understand de Man to be saying that his- tory is caused by "intentional" uses of language and that might therefore mislead you, dear reader. As any careful reader of de Man knows, his theory of the perfor- mative "use" of language (as opposed to its mention) is detached from any con- scious intention in the user. Language works performatively, on its own, most often against the intentions or knowledge of the speaker or writer. As he says, in the con- clusion to "Promises (Social Contract)," "The error is not within the reader; lan- guage itself dissociates the cognition from the act. Die Sprache verspricht (sich)" (AR 277), which means "Language promises" and also "Language makes a slip of the tongue. " I have thought it worthwhile to refer directly to the comments of the Minnesota reader in order to try to forestall similar errors on the part of readers of the published essay.
8. Louis Althusser, "Ideology and Ideological State Apparatuses (Notes to- wards an Investigation," in Lenin and Philosophy and Other Essays, trans. Ben Brewster (New York: Monthly Review Press, 1972), 162. See page 159, where Althusser says, "ideology has no history," and goes on to remark: "As we know, this formulation appears in so many words in a passage from The German Ideology. "
9. In an equally important, though much less well known, definition of ideolo- gies near the beginning of "Phenomenality and Materiality in Kant" de Man assert- ed that ideologies are on the side of what Kant called "metaphysics," that is, in Kant's use of the term, precritical empirical knowledge of the world. Only critical analysis of ideologies will keep ideologies from becoming mere illusion and critical philosophy from becoming idealism cut off from the empirical world (AI 72). The anonymous reader for the University of Minnesota Press sternly challenged my understanding in this footnote of Kant's use of the term metaphysics. This is another allergic reaction, one that demonstrates just the point I am making about de Man. Surely Kant cannot have meant something so strange as this by "metaphysics"! At the risk of making this footnote tediously long for those who have read Kant and de Man's commentary on Kant, here is the relevant passage from Kant, followed by de Man's comment on it. I think my reader is mystified through having accepted received opinion about what Kant must be saying because everyone knows that is what he says. That received opinion is, precisely, a species of "ideology," even of
"aesthetic ideology. " Kant says: "A transcendental principle is one through which we represent a priori the universal condition under which alone things can become Objects of our cognition generally. A principle, on the other hand, is called meta- physical [Dagegen heisst ein Prinzip metaphysisch], where it represents a priori the condition under which alone Objects whose concept has to be given empirically [empirisch], may become further determined [bestimmet] a priori. Thus the prin- ciple of the cognition of bodies [der Erkenntnis der Ko? rper] as substances, and as changeable substances, is transcendental where the statement is that their change must have a cause [Ursache]: but it is metaphysical where it asserts that their change must have an external cause [eine a? ussere Ursache]. For in the first case bod- ies need only be thought through ontological predicates (pure concepts of under- standing [reine Verstandesbegriffe]), e. g. as substance, to enable the proposition to be cognized a priori; whereas, in the second case, the empirical concept of a body (as a movable thing in space) must be introduced to support the proposition [diesem Satze zum Grunde gelegt werden muss], although, once this is done, it may be seen [eingesehen] quite a priori that the latter predicate (movement only by means of an external cause) applies to body" (Immanuel Kant, Kritik der Urteilskraft, ed. Wilhelm Wieschedel [Frankfurt am Main: Suhrkamp, 1979], 90; The Critique of Judgement, trans. James Creed Meredith [Oxford: Oxford University Press, 1982], 20-21). De Man comments, in "Phenomenality and Materiality in Kant": "The condition of existence of bodies is called substance; to state that substance is the cause of the motion of bodies (as Kant does in the passage quoted) is to exam- ine critically the possibility of their existence. Metaphysical principles, on the other hand, take the existence of their object for granted as empirical fact. They contain knowledge of the world, but this knowledge is precritical. Transcendental prin- ciples contain no knowledge of the world or anything else, except for the knowl- edge that metaphysical principles that take them for their object are themselves in need of critical analysis, since they take for granted an objectivity that, for the tran- scendental principles, is not a priori available. Thus the objects of transcendental principles are always critical judgments that take metaphysical knowledge for their target. Transcendental philosophy is always the critical philosophy of metaphysics" (AI 71). De Man goes on to associate ideology with metaphysics as Kant defines it. The passage is an important gloss on de Man's definition, or, more properly, "call- ing," of ideology in "The Resistance to Theory," just cited. In the sentences that follow just after the ones already quoted from "Phenomenality and Materiality in Kant" de Man associates ideology with Kantian "metaphysics" and argues for an intricate interdependence of critical thought on ideology and of ideology, if it is to other than "mere error," on critical thought.
